% !BIB TS-program = biber
\documentclass[a4paper,12pt]{article}

\usepackage[utf8]{inputenc}
\usepackage{etex}
\usepackage{mathtools}
\usepackage{verbatim}
\usepackage{amssymb,amsmath,epsfig,rotating,ushort}
\usepackage{amsthm}
\usepackage{gensymb}
\usepackage{thmtools}
\usepackage{setspace}
\usepackage{a4wide}
\usepackage{a4}
\usepackage[labelfont=bf,font=normalsize]{caption}
\usepackage{tabularx}
\usepackage{array}
\usepackage{rotating}
\usepackage{pdflscape}
\usepackage{chngpage}
\usepackage{textcomp}
\usepackage{color, fancybox}
\usepackage{amsthm, amstext}
\usepackage{mdwlist}
\usepackage{threeparttable}
\usepackage{setspace}
\usepackage{float,graphicx, color}
\usepackage{lscape}
\usepackage[marginal,stable]{footmisc}
\usepackage{fancyhdr}
%\usepackage{pst-all,pstricks-add}
%\usepackage{sectsty}
%\usepackage{subfigure}
%\usepackage[round]{natbib}
%\usepackage{cite}
\usepackage{hyperref}
\usepackage{caption}
\usepackage{subcaption}
\usepackage{booktabs,setspace}
\usepackage{footmisc}
\usepackage{sectsty}
%\usepackage{titlesec}
\usepackage[maxcitenames=2,mincitenames=1,natbib,style=authoryear,backend=biber,url=false,doi=false,eprint=false,isbn=false,uniquelist=false]{biblatex}

\usepackage{titling}
\setlength{\droptitle}{-0.5in}


\addbibresource{literature.bib}
\usepackage[normalem]{ulem}
\usepackage[margin=2.54cm]{geometry}
\AtEveryBibitem{\clearfield{month}}
\AtEveryCitekey{\clearfield{month}}
\renewbibmacro{in:}{}

\renewcommand{\arraystretch}{.75}
%\addtolength{\skip\footins}{10mm}
\setlength{\footnotesep}{13pt}
%\setlength{\oddsidemargin}{-0.3cm}
%\setlength{\topmargin}{1.3cm}
\addtolength{\topmargin}{1.27cm}
%\setlength{\textwidth}{16.5cm}
\addtolength{\textheight}{-2.54cm}

\newcommand{\sym}[1]{$^{#1}$}
\newcolumntype{H}{>{\setbox0=\hbox\bgroup}c<{\egroup}@{}}

\renewcommand{\arraystretch}{1.2}


%\renewcommand{\thesection}{A.\arabic{section}}
%\renewcommand{\theequation}{A.\arabic{equation}}
%\renewcommand{\thefigure}{A.\arabic{figure}}
%\renewcommand{\thetable}{A.\arabic{table}}

\paragraphfont{\fontsize{11}{1em}\selectfont}
\sectionfont{\fontsize{14}{1em}\selectfont}
\subsectionfont{\fontsize{13}{1em}\selectfont\itshape\bfseries}
\subsectionfont{\fontsize{13}{1em}\selectfont\bfseries}
\paragraphfont{\fontsize{11}{1em}\selectfont}
\renewcommand{\baselinestretch}{1.2}\normalsize


\doublespacing


\begin{document}



\makeatletter\providecommand{\tabularnewline}{\\}\makeatother
\renewcommand{\tabcolsep}{.4cm}




\begin{table}[t!]\centering
\caption{Public and private firms (2016)}\label{tab:firms}
\footnotesize
\begin{tabular}{lccccc}
&&& & share of sectors & share of firms with  \\
& \# firms & share  & revenue share &   with \# firms $\ge 2$ & common ownership \\
\midrule
\input{tab1}\\[-1em]
\bottomrule
        \multicolumn{6}{p{.95\textwidth}}{\scriptsize This table presents descriptive statistics. The sample includes all non-financial joint-stock firms in Norway in 2016. Public firms include listed firms and firms owned by them. Sector corresponds to 5-digit level code of NACE Rev. 2. \# sectors: 742. Firms with common ownership are firms that have a household blockholder who also holds a block in another firm in the same sector. Blockholdings are defined as ultimate shareholdings $> 5\%$.} \\
\end{tabular}{\small\par}
\end{table}


\renewcommand{\tabcolsep}{.1cm}
\begin{table}[h]
\caption{Ownership of public and private firms}\label{tab:opf}
\footnotesize
\begin{tabular}{llccccccccccc}
& \multicolumn{5}{c}{\textit{all shareholders}} & \multicolumn{5}{c}{\textit{blockholders}} \\
\cmidrule(lr){2-6}\cmidrule(lr){7-11}
%\cmidrule(lr){2-3}\cmidrule(lr){7-8}
& direct & \multicolumn{4}{c}{ultimate} & direct & \multicolumn{4}{c}{ultimate}\\
\cmidrule(lr){2-2} \cmidrule(lr){3-6}\cmidrule(lr){7-7}\cmidrule(lr){8-11}
\# shareholders: &  \multicolumn{1}{c}{total} & total & individ. &  institutions & foreign&  \multicolumn{1}{c}{total} & total & individ. &  institutions & foreign&\\
\midrule\\[.2em]
\multicolumn{4}{l}{\textit{public firms}} \\
\cmidrule(r){1-1}
\input{tab2}\\
\multicolumn{4}{l}{\textit{private firms}} \\
\cmidrule(r){1-1}
\input{tab2_2}\\
\bottomrule
        \multicolumn{11}{p{1\textwidth}}{\scriptsize This table presents descriptive statistics. The sample includes all non-financial joint-stock firms in Norway in 2016. Public firms include listed firms and firms owned by them. The number of public firms is 3,603, the number of private firms is 192,984. Blockholdings are defined as shareholdings $> 5\%$.} \\
\end{tabular}{\small\par}
\end{table}






\renewcommand{\tabcolsep}{.5cm}

\begin{table}[h!]\centering
\caption{Composition of household blockholders' portfolios}\label{tab:fpo}
\footnotesize
\begin{tabular}{lcccccc}
& mean & p(50) & p(90) & p(95) & interquartile \\
\textit{all stocks} &&&&& range\\
\midrule
\input{tab3}\\
\multicolumn{4}{l}{\textit{blockholdings}} \\
\midrule
\input{tab3_2}\\[-1em]
\bottomrule
        \multicolumn{6}{p{.78\textwidth}}{\scriptsize This table shows summary statistics for the population of 220,716 Norwegian household shareholders in 2016 that have at least one blockholding (ultimate ownership share $> 5\%$) in a private non-financial firm.} \\
\end{tabular}{\small\par}
\end{table}


\renewcommand{\tabcolsep}{.75cm}


\begin{figure}[t!]\centering
\caption{Distribution of blockholdings per owner}\label{fig:blockholdings}
\includegraphics[scale=.8]{fig1.pdf}
\begin{minipage}{11cm}
\scriptsize This figure shows the share of households that block-hold a given number of firms. The sample consists of the population of 220,716 Norwegian household shareholders in 2016 that have at least one blockholding (ultimate ownership share $> 5\%$) in a private non-financial firm.
\end{minipage}
\end{figure}


\begin{figure}[h!]\centering
\caption{Diversification of blockholdings}\label{fig:diversification}
\includegraphics[scale=.8]{fig2.pdf}
\begin{minipage}{11cm}
\scriptsize This figure shows the average (grey squares) and minimum (red dots) number of sectors spanned by the blockholdings by households with a given number of block held firms. The sample consists of the population of 220,716 Norwegian household shareholders in 2016 that have at least one blockholding (ultimate ownership share $> 5\%$) in a private firm. Bins with less than five households are not shown.
\end{minipage}
\end{figure}

 
 
\begin{figure}[t]\centering
\caption{Common ownership across aggregate sectors}\label{fig:cowned}
\includegraphics[scale=.8]{fig3.pdf}

\begin{minipage}{11cm}
\scriptsize This figure shows the share of firms with common ownerships in the four aggregate sectors of the economy: agriculture, manufacturing, mining, and services, for both private and public firms. Firms with common ownership are firms that have a household blockholder who also holds a block in another firm in the same 5-digit sector of the NACE Rev 2 classification. Blockholdings are defined as ultimate shareholdings $> 5\%$. The sample includes all non-financial joint-stock firms in Norway in 2016. The number of public firms is 3,603, the number of private firms is 192,984. Public firms include listed firms and firms owned by them.
\end{minipage}
\end{figure}


\begin{figure}\centering
\caption{HHI and $\Delta$ MHHI}\label{fig:mhhi}
\includegraphics[scale=.4]{fig4.png}
\begin{minipage}{13cm}
\scriptsize This figure shows the 40 sectors  (5-digit categories of NACE Rev. 2 classification) with the highest degree of common ownership as measured by $\Delta MHHI$, calculated using blockholdings  (ultimate ownership share $> 5\%$) of households and foreign owners (dark blue bars). The figure also shows the normalized $HHI$ for these sectors (grey bars) and the contribution of blockholdings by other ultimate owners (e.g. government institutions) to $\Delta MHHI$ (light blue bars).  The sample consists of all non-non-financial joint-stock firms in 2016 (196,590 firms).
\end{minipage}
\end{figure}


\begin{figure}[t]\centering
\caption{MHHI based on ultimate vs. direct ownership}\label{fig:mhhi_dir_ult}
\includegraphics[scale=.35]{fig5.png}
\begin{minipage}{11cm}
\scriptsize This figure plots $\Delta MHHI$ calculated using ultimate ownership shares against $\Delta MHHI$ calculated using direct ownership shares. Only blockholdings (direct or ultimate ownership shares $>5\%$) by households and foreign owners are considered. Each dot corresponds to a sector (5-digit category of NACE Rev. 2 classification) The sample consists of all non-non-financial joint-stock firms in 2016 (196,590 firms).
\end{minipage}
\end{figure}
\renewcommand{\tabcolsep}{.2cm}

\begin{table}[t!]\centering
\caption{Overlap of blockholdings and corporate roles}\label{tab:overlap}
\footnotesize
\begin{tabular}{lccccc}
&\multicolumn{1}{c}{share of firms} & share of blockholders \\
& in which blockholder is & that are  \\
\midrule
\input{tab4}\\[-1em]
\bottomrule
        \multicolumn{3}{p{.9\textwidth}}{\scriptsize This table presents descriptive statistics on the overlap of blockholdings and corporate roles. Column 1 shows the share of firms in which at least one blockholder is also the CEO or an independent board member, and the share of firms in which this blockholder also holds a block in a competitor firm. A competitor firm is defined as a firm in the same NACE Rev. 2 5-digit category. Column 2 shows the share of blockholders that are also CEOs or independent board members in at least one of their portfolio firms and the share of these blockholders that also hold a block in a competitor firm. The sample consists of all non-non-financial joint-stock firms in 2016 (196,590 firms).} \\
\end{tabular}{\small\par}
\end{table}
 


\renewcommand{\tabcolsep}{.4cm}


\vspace*{0cm}\begin{table}[t!]
	\centering
	\scriptsize
\caption{ Balancing test: Within marriages}
	\label{tab:balance_within}
   		\begin{tabular}{lcccccccc}
&  \multicolumn{4}{c}{unconditional}  &  \multicolumn{2}{c}{conditional}   \\ 
\cmidrule(lr){2-5}\cmidrule(lr){6-7}
\input{tab5}\\[-1em]
        		\midrule
        \multicolumn{7}{p{.9\textwidth}}{\scriptsize This table presents balancing tests for firm characteristics of treated and control firms in the within-marriage setting. Columns titled ``treatment" and ``control" show pre-treatment averages for the respective groups of firms. Columns titled $\Delta$ $\Delta$ show the pre-treatment mean difference between treated and controls. Columns titled $p(\Delta=0)$ shows the p-value of the test for $\Delta=0$. The columns titled "conditional" compare means conditional on 2-digit-NACE-sector$\times$year fixed effects. Sample consists of six-year periods around the event time, within the period 2005 -- 2018.  Operating income, wage bill, share capital, and dividend are in USD.}  \\
        \bottomrule
    \end{tabular}
\end{table}


\vspace*{0cm}\begin{table}[h!]
	\centering
	\scriptsize
\caption{ Balancing test: Within sectors}
	\label{tab:balance_across}
   		\begin{tabular}{lcccccccc}
&  \multicolumn{4}{c}{unconditional}  &  \multicolumn{2}{c}{conditional}   \\ 
\cmidrule(lr){2-5}\cmidrule(lr){6-7}
\input{tab6}\\[-1em]
        		\midrule
        \multicolumn{7}{p{.9\textwidth}}{\scriptsize This table presents balancing tests for firm characteristics of treated and control firms in the within-sector setting. Columns titled ``treatment" and ``control" show pre-treatment averages for the respective groups of firms. Columns titled $\Delta$ show the pre-treatment mean difference between treated and controls. Columns titled $p(\Delta=0)$ shows the p-value of the test for $\Delta=0$. The columns titled "conditional" compare means conditional on 2-digit-NACE-sector$\times$year fixed effects. Sample consists of six-year periods around the event time, within the period 2005 -- 2018.  Operating income, wage bill, share capital, and dividend are in USD.}  \\
        \bottomrule
    \end{tabular}
\end{table}



\begin{figure}[ht]
  \centering
    \caption{Event study analysis: Effect on profit margins}
  \begin{subfigure}[b]{0.45\textwidth}
    \centering
    \includegraphics[width=\textwidth]{fig6a.pdf}
    \caption{Within marriages}
    \label{fig:subfig1}
  \end{subfigure}
  \hfill
  \begin{subfigure}[b]{0.45\textwidth}
    \centering
    \includegraphics[width=\textwidth]{fig6b.pdf}
    \caption{Within sectors}
    \label{fig:subfig2}
  \end{subfigure}
  \label{fig:event}
\begin{minipage}{\textwidth}
\scriptsize Panels (a) and (b) of this figure plot event study coefficients $\beta_{\tau}$ from in Eq. \eqref{eq:WM} (within-marriage setting)  and Eq. \eqref{eq:AM} (within-sector setting), respectively. The event-study coefficients measure the difference between the profit margins of treated and control firms before, during and after the treatment year (period 0), conditional on fixed effects. Profit margins are measured by the ratio of profits before tax to operating income. Grey bars indicate 90\% confidence intervals. Standard errors clustered on sector (5-digit level of NACE Rev. 2. Panel (a) is based on 135 marriages, 352 treated firms and 455 control firms. Panel (b) is based on 1,454 marriages, 545 treated firms and 2,838 control firms. Treatment indicator is constructed using a 5\% threshold for blockholdings. Sample period: 2005-2018.
\end{minipage}
\end{figure}


\renewcommand{\tabcolsep}{.6cm}
\begin{table}[h]
	\centering
	\scriptsize
	\caption{Event study analysis: Regression results}
	\label{tab:event05}
    %
		\begin{tabular}{lcccccccc}
				 \multicolumn{3}{l}{Dependent variable: $ProfitMargin_{j,t}$} \\
		\toprule

	%	& (1) & (2) & (3)  & (4)  & (5)  \\[.1em]
		&   \multicolumn{2}{c}{Within marriages }  &  \multicolumn{2}{c}{Within sectors } \\ 
\cmidrule(lr){2-3}\cmidrule(lr){4-5}\\[-1em]
& (1) & (2) & (3) & (4)  \\
		\midrule
\input{tab7}\\[.3em]
        		\midrule
        Fixed effects & $j,mt,s_2t$ & $j,mt,s_2t$ & $j,s_5t$& $j,s_5t$ \\ 
\input{tab7_base}\\[-1em]
        \midrule

        \multicolumn{5}{p{.74\textwidth}}{\scriptsize This table reports estimated event-study coefficients, which measure the difference between the profit margins of treated and control firms before, during and after the treatment year. Columns 1,2 and 3,4 show the estimated event-study coefficients $\beta_{\tau}$ in Eq. \eqref{eq:WM} (within-marriage setting)  and \eqref{eq:AM} (within-sector setting), respectively. Columns 1,2 include firm, marriage-event-time, and sector (2-digit)-year fixed effects. Columns 3,4 include firm, and sector (5-digit)-year fixed effects. Standard errors clustered on sector (5-digit level of NACE Rev. 2). $^{***}p<0.01$, $^{**}p<0.05$, $^*p<0.1$. $T_{t+1}$ is the omitted category. $Post$ is an indicator for post treatment periods ($\tau\geq 0$). Columns 1,2 are based on 135 marriages, 352 treated firms and 455 control firms. Columns 3,4 are based on 1,454 marriages, 545 treated firms and 2,838 control firms. Treatment indicator is constructed using a 5\% threshold for blockholdings. Sample period: 2005-2018.}  \\
        \bottomrule
    \end{tabular}
\end{table}



\begin{figure}[h!]
  \centering
    \caption{Event study analysis: Effect on profit margins (threshold 10\%)}
  \begin{subfigure}[b]{0.45\textwidth}
    \centering
    \includegraphics[width=\textwidth]{fig7a.pdf}
    \caption{Within marriages}
    \label{fig:subfig1}
  \end{subfigure}
  \hfill
  \begin{subfigure}[b]{0.45\textwidth}
    \centering
    \includegraphics[width=\textwidth]{fig7b.pdf}
    \caption{Within sectors}
    \label{fig:subfig2}
  \end{subfigure}
  \label{fig:event1}
\begin{minipage}{\textwidth}
\scriptsize Panels (a) and (b) of this figure plot event study coefficients $\beta_{\tau}$ from in Eq. \eqref{eq:WM} (within-marriage setting)  and Eq. \eqref{eq:AM} (within-sector setting), respectively. The event-study coefficients measure the difference between the profit margins of treated and control firms before, during and after the treatment year (period 0), conditional on fixed effects. Profit margins are measured by the ratio of profits before tax to operating income. Grey bars indicate 90\% confidence intervals. Standard errors clustered on sector (5-digit level of NACE Rev. 2. Panel (a) is based on 135 marriages, 352 treated firms and 455 control firms. Panel (b) is based on 1,454 marriages, 545 treated firms and 2,838 control firms. Treatment indicator is constructed using a 10\% threshold for blockholdings. Sample period: 2005-2018.
\end{minipage}
\end{figure}



\renewcommand{\tabcolsep}{.2cm}
\begin{table}[t!]
	\centering
	\scriptsize
	\caption{Profit margins and common ownership}
	\label{tab:main}
    %
		\begin{tabular}{lcccccccc}
				 \multicolumn{3}{l}{Dependent variable: $ProfitMargin_{j,t}$} \\
		\toprule

	%	& (1) & (2) & (3)  & (4)  & (5)  \\[.1em]
		&   \multicolumn{4}{c}{Ownership share $>.05$ }  &  \multicolumn{4}{c}{Ownership share $>.1$ } \\ 
\cmidrule(lr){2-5}\cmidrule(lr){6-9}\\[-1em]
		& OLS & OLS & 2SLS & 2SLS & OLS & OLS & 2SLS & 2SLS\\
& (1) & (2) & (3) & (4) & (5) & (6) & (7) & (8) \\
		\midrule
\input{tab8}\\[.3em]
        				 \multicolumn{3}{l}{\textit{First-stage:}} 				 \\
        \midrule
%$\mathcal{I}_{marriage}$  & & &  .076$^{***}$&  .187$^{***}$  & &  &.119$^{***}$ &.291$^{***}$ \\
%& & &(.029) & (.070)  && & (.035)  & (.093)  \\
        \midrule
   %     F-stat & & &  6.81 & 7.02  & && 11.67 & 9.7  \\
        		\midrule
        Fixed effects & $i,st$ & $i,st$ & $i,st$& $i,st$& $i,st$ & $i,st$ & $i,st$ & $i,st$  \\ 
\input{tab8_base}\\[-1em]
        \midrule

        \multicolumn{9}{p{.9\textwidth}}{\scriptsize This table presents estimates of the effect of common ownership on firm-level profit margins from OLS and 2SLS regressions. $CrossDummy$ is a binary indicator equal to one if a firm has a blockholder in common with another firm in the same sector (5-digit level of NACE Rev. 2). $TotalCrossOwn$ is a continuous measure of firm-level common ownership. In columns 1-4 (5-8) common ownership measures are calculated with a cut-off for blockholdings for 5\% (10\%). In the 2SLS regressions (columns 3,4,7,8), the excluded variable is $\mathcal{I}_{marriage}$, a dummy equal to one if a blockholder of the firm gets married to someone who is a blockholder in one or more firms in the same sector. Standard errors clustered on sector. Firm fixed effects and sector$\times$year fixed effects included.$^{***}p<0.01$, $^{**}p<0.05$, $^*p<0.1$.  F-stat is the Kleinbergen-Paap F-statistic. The sample includes all non-financial joint stock firms in Norway over the period 2005 -- 2018.}  \\
        \bottomrule
    \end{tabular}
\end{table}




\renewcommand{\tabcolsep}{.2cm}
\begin{table}[h]
	\centering
	\scriptsize
	\caption{Event study analysis: Other outcome variables}	\label{tab:other_out}
		\begin{tabular}{lcccccc}
				\toprule
	%	& (1) & (2) & (3)  & (4)  & (5)  \\[.1em]
		&   \multicolumn{3}{c}{Within marriages }  &  \multicolumn{3}{c}{Within sectors } \\ 
\cmidrule(lr){2-4}\cmidrule(lr){5-7}\\[-1em]
& (1) & (2) & (3) & (4) & (5) & (6)  \\
Dep. var.: & $\ln$ OperatingIncome   & $\ln$ Sales   & $\frac{R\&D}{Assets}$ & $\ln$ OperatingIncome  & $\ln$ Sales   & $\frac{R\&D}{Assets}$ \\
		\midrule
\input{tab9}\\[.3em]
        		\midrule
        Fixed effects & $j,mt,s_2t$ & $j,mt,s_2t$  & $j,mt,s_2t$ & $j,s_5t$ & $j,s_5t$& $j,s_1t$ \\ 
\input{tab9_base}\\[-1em]
        \midrule

        \multicolumn{7}{p{.86\textwidth}}{\scriptsize This table presents estimates of the effect of common ownership on operating income (in logs) and on R\&D expenditure (measured as share in total assets). Columns 1,2 (3,4) use the treated and control firms from the within-marriage (within-sector) event study setting, comprising 135 marriages events, 352 treated firms and 455 control firms (1,454 marriages events, 545 treated firms and 2,838). Columns 1,2 include firm, marriage-event-time, and sector (2-digit)-year fixed effects. Column 3 include firm and sector (5-digit)-year fixed effects. Column 4 includes firm and sector (1-digit)-year fixed effects. 
        Standard errors clustered on sector (5-digit level of NACE Rev. 2). $^{***}p<0.01$, $^{**}p<0.05$, $^*p<0.1$. $Post$ is an indicator for post treatment periods ($\tau\geq 0$).}  \\
        \bottomrule
    \end{tabular}
\end{table}
 

\end{document}
